Browsing by Author "Jahidul Islam, Md."

Filter results by typing the first few letters
Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    Improving Bangla Linguistics: Advanced LSTM, Bi-LSTM, and Seq2Seq Models for Translating Sylheti to Modern Bangla
    (2024-04-04) Das, Sourav Kumar; Naeen, Md. Julkar; Jahidul Islam, Md.; Sajeeb, Md. Anisul Haque; Chakraborty, Narayan Ranjan; Mojumdar, Mayen Uddin
    Bangla or Bengali is the national language of Bangladesh, people from different regions don’t talk in proper Bangla. Every division of Bangladesh has its own local language like Sylheti, Chittagong etc. In recent years some papers were published on Bangla language like sentiment analysis, fake news detection and classifications, but a few of them were on Bangla languages. This research is for the local language and this particular paper is on Sylheti language. It presented a comprehensive system using Natural Language Processing or NLP techniques for translating Pure or Modern Bangla to locally spoken Sylheti Bangla language. Total 1200 data used for training 3 models LSTM, Bi-LSTM and Seq2Seq and LSTM scored the best in performance with 89.3% accuracy. The findings of this research may contribute to the growth of Bangla NLP researchers for future more advanced innovations.
  • Thumbnail Image
    Item
    Intelligent dynamic spectrum access exploiting a synergy between genetic algorithm and local search
    (Department of Computer Science and Engineering, 2015-02) Jahidul Islam, Md.; Monirul Islam, Dr. Md.
    This thesis presents a novel hybrid dynamic spectrum access technique for multi-channel single-radio cognitive radio networks. Existing classical and stochastic approaches exhibit di erent advantages and disadvantages depending on network topology and architecture. Our proposed approach exploits a delicate balance between these two types of approaches for extracting advantages from both of them while limiting their disadvantages. We exploit a synergy between genetic algorithm-based stochastic search and classical local search to design a highly scalable and e cient dynamic spectrum access technique. Additionally, we boost up the performance of our algorithm through designing new genetic operators. Besides, proper and thorough performance evaluation of existing approaches using a discrete event simulator is yet to be performed in the literature. To address this issue, we simulate several existing approaches using a widely used discrete event simulator called ns-2. We evaluate the performance of our proposed technique in ns-2 on the basis of various standard performance metrics. In the evaluation, we compare the performance of our proposed technique with that of the state-of-the-art approaches. Simulation results demonstrate signi cant performance improvement using our proposed approach over the existing ones.